Accuracy of 2-D versus 3-D ultrasound in measurement of cervical length

Abstract

OBJECTIVE: Validation and reliability of 2-D vs. 3-D ultrasound measurement of cervical length (CL) and comparison to the gold-standard (3-D ultrasound with a balloon catheter at the internal os) DESIGN: Validation study comparing 2-D and 3-D ultrasound CL measurements MATERIALS AND METHODS: 20 non-gravid volunteers, ages 18 to 39, had mid-follicular ultrasounds for CL. Three 2-D and three 3-D measurements were recorded by a single ultrasonographer who was blinded to the results. A balloon catheter was then placed at the internal os and the measurements were repeated. Data were compared using Pearson correlation coefficients and paired t-tests. RESULTS: Internal reliability of measurements were analyzed using Pearson correlations (2-D, 2-D with balloon, 3-D, and 3-D with balloon). The 1st measurement in each set was discarded, and the 2nd and 3rd were compared. All correlation coefficients were 0.80 or higher. Measurements were validated by comparing without balloon measurements to with balloon measurements.Tabled 1Comparison of Differences in Cervical Length by Various Ultrasound MeasurementsMeasurementDifference in Cervical LengthSD95% Confidence Intervalp value2-D vs. 2-D with balloon0.510.620.22, 0.800.0023-D vs. 3-D with balloon0.350.690.03, 0.670.042-D vs. 3-D with balloon0.580.680.26, 0.900.001 Open table in a new tab Without balloon measures significantly overestimated CL when compared to with balloon measure. When compared to the gold standard, 2-D and 3-D significantly overestimated CL (0.58 cm and 0.35 cm respectively), but 3-D provided a better approximated of the gold-standard. CONCLUSIONS: All measurment techniques were internally reliable in the hands of an experienced ultrasonographer. Both without balloon measurement overestimated cervical length in the non-gravid uterus, but 3-D measurements better approximated the “gold standard” CL than 2-D measurements.